TOPIC: Speaker Impedance

Hi Guys,

I want to measure the input voltage that makes my amplifier clip so that I can reduce the gain on the 4 output channels and ensure that MV0 is below the amp clipping point.

I don’t want speakers attached to the amp when driving it to clipping so I understand I’ll need to create a dummy load for the amplifier.
I know the frequency at which my (4 Ohm nominal) speakers produce their worst load (which is 3 Ohms).
I’ll get a 50W 3 Ohm resistor for the load.
I’ll use REW to produce the sine wave.
I’ll use a basic oscilloscope to monitor the sine wave for clipping.
This will test the amp in the worst case scenario.

But music isn’t a single tone and the worst case scenario isn’t going to occur very often so I’m interested to see where clipping occurs whilst playing pink noise.

What value resistor should I use for this test? Can I somehow (at lower volume) measure the load my speakers produce when playing pink noise?

Thanks

Can you explain why you are worried about clipping your amplifier? It is not uncommon to have your gain structure set up such that you can clip your amplifier at a high volume as this allows you to compensate for lower level recordings and as you mention music is not a continuous signal.

How do you propose to see when the pink noise is clipping? It is a random signal and it will be difficult to accurately see the exact clipping point (although you will be able to see gross clipping).

I would just stick with a sine wave. The resistor should be sized such that it can handle the full output of your amplifier.

Michael

mdsimon2 wrote:
Hmmm yes.. fair point.

My concern only started when I went from a solid state amp with 100s of watts to mono block valve power amps which are only rated 30w. My speaker sensitivity is 87db/w/m and although the combination is loud enough at my listening position I keep reading about music having dynamics that could be plus 20db above listening level. That makes me wonder when these peaks occur, if my amp is clipping or not.

I am not familiar with valve amps but isn't part of the magic that they clip nicely? .

Do the amps have a specified input sensitivity? I know we had talked about measuring gain before as well, if you measured the gain and felt like you could trust the 30W rating you could also work out the input sensitivity and compare that to your maximum line level voltage.

Michael

Hi again @mdsimon2, I’m not sure the ‘soft clipping’ is part of the magic, it’s more of a bonus! These are the first valve amps I’ve ever had and they replaced the very well thought of Krell KAV 400xi.

The valves probably dig out as much detail as the Krell did but what I immediately noticed as an improvement (which I would say is part of the magic) was that the valve amps sounded less harsh and brittle and more natural. Listening sessions never get tiring in fact it’s only having to go to bed that ends them!

The trade off is power. These amps are from the early 90s and when I had them recapped the power was reduced a small amount to keep modern (lower quality) vacuum tubes alive longer.

I therefore can’t say for sure that they are still 30w per channel. Input sensitivity I’m unsure of too. The manual says it’s 0.45v but a magazine review says they measured it as 0.3v.

With regards to gain, I know you have previously given me instructions how to measure this. I do still mean to and appreciated your guidance. Just with a young child and everything going on with lockdowns, finding the time to take the measurements is a bit tricky but it’s on my to do list!!

Once I’ve measured/worked out the gain, if I was to know the input sensitivity could I work out the new power rating?

Also what could I calculate if I knew the input voltage that causes clipping?

Thanks

The gain really only helps if you know the power rating. If you know the gain and the power rating you could back calculate input sensitivity.

If you know the input sensitivity you can determine above what volume level you may get clipping. I say may because music is dynamic and rarely reaches 0 dB so just because you can clip the amp doesn't mean you actually will and whether you will clip or not depends on how the music was recorded. It also depends how much boost you have in the DDRC and the music level at the boost frequencies.

The input sensitivities you mentioned and a 30W power rating at 8 ohms imply a gain between 31 and 34 dB. At 0.45V input sensitivity you could get clipping at a volume levels higher than -14 dB and at 0.3V input sensitivity you could get clipping at volume levels higher than -17 dB.

All that being said if you are happy with the sound I am not sure that it really matters if you are clipping the amp during high level transients.

Michael

mdsimon2 wrote:
Hi @mdsimon2,
Could you show the calculation you did for the above? The 30W is at 6Ohms so would like to understand your calculation so I can try a couple of different combination of values.

Thanks

P = V^2 / R -> V = sqrt(P*R)
G = 20*log(Vout/Vin)

For demonstration using my previous numbers for 30W @ 8 ohms, 0.45V input sensitivity and DDRC output voltage of 2.08V.

Amp Clipping Voltage = sqrt(30*8) = 15.5V
Amp Gain = 20*log(15.5/0.45) = 30.7 dB
Clipping Volume = 20*log(0.45/2.08) = -13.3 dB

As you can see the clipping volume only depends on input sensitivity and DAC output voltage.

Using 30W @ 6 ohms:

Amp Clipping Voltage = sqrt(30*6) = 13.4V
Amp Gain = 20*log(13.4/0.45) = 29.5 dB

Michael

@mdsimon2,

Thanks for that - much appreciated.

What is the difference between Amp Clipping Voltage and Clipping Volume? Is it that once I've raised the DDRC's master volume to the level that will cause the amp to clip (Clipping Volume), the amplifier's output voltage will equal/be the Clipping Voltage?

Your understanding is correct.

The amp clipping voltage is the output voltage from the amplifier at its power rating. By definition you will reach this amplifier output voltage when you provide the amplifier an input voltage equivalent to the input sensitivity.

Michael

Great....

Now I’m not sure I dare ask but how does current come in to play? If my speakers drop to 3ohms at a certain frequency, the voltage doesn’t change but the current does right.

Is it now a case of V=IR ?
13.4 /6 = 2.2 amps max is it?

13.4 / 3 = 4.47
What are the implications of that?

I guess the question is whether the power supply can keep up and supply that current. I am not an EE and am very unfamiliar with tube amps but my understanding is that they will handle this situation more gracefully than solid state amps and although both will be driven to higher levels of distortion the tube amp will not hard clip like a solid state amp.

Michael

Hi Again, Sorry....

How do you work out the correlation between DDRC Vout and Amplifier Vout or Watts?

So if I know for example that for SPL 76db at the listening position I need 0.17W, and at 6 Ohm this would be 1v out from the Amp. How do you calculate the corresponding Vout at the DDRC?

Thanks

The amplifier gain is what correlates the DDRC output voltage and the amplifier output voltage. As we discussed we really do not know what the gain is at this point but from the input sensitivity and power rating we can estimate it.

G = 20*log(Vout/Vin) -> rearranging to solve for Vin -> Vin = Vout*10^(-G/20) [this is the DDRC output voltage for a given amplifier output voltage]

We think the amplifier gain is somewhere between 29.5 dB (0.45V input sensitivity, 30W @ 6 ohm) and 33.0 (0.30V input sensitivity, 30W @ 6 ohm). I've calculated the DDRC output voltage required to reach 1V amplifier output voltage for each case below.

0.45V input sensitivity
Vin = 1*10^(-29.5/20) = 0.0335V
DDRC Volume = 20*log(0.0335/2.08) = -35.9 dB

0.30V input sensitivity
Vin = 1*10^(-33.0/20) = 0.0224V
DDRC Volume = 20*log(0.0224/2.08) =-39.4 dB

All of this is a bit of simplification as these voltages assume 0 dB input signals which is almost never the case.

Michael

Brilliantly helpful again. Thanks. And yes, take your point about music input signal never really reaching 0db.